Tutorograd commited on
Commit
6b9bd38
1 Parent(s): 7667e3e

Create koikatsu_animation_workflow.md

Browse files
Files changed (1) hide show
  1. koikatsu_animation_workflow.md +47 -0
koikatsu_animation_workflow.md ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ This is my second attempt in my quest for consistent animation optimization that I thought it was worth to share this time.
2
+
3
+ It directly uses computed depth frames from a 3D motion here, which means clean depth, allowing qualitative character swap. This approach is different from real-to-anime img2img chick videos. So there is no video reference. Good thing is it avoids the EBSynth hassle. Also VERY few manual aberration correction.
4
+
5
+ The workflow is a bit special since it uses the Koikatsu h-game studio. I guess Blender works too. But this "studio" is perfect for 3D character and pose/scene customization with awesome community and plugins (like depth). The truth is I have more skills in Koikatsu than in Blender (shhh~).
6
+
7
+ Here is the workflow, and I probably need some advice from you to optimize it:
8
+
9
+ KOIKATSU STUDIO
10
+
11
+ Once satisfied with the motion (can be MMD), extract the depth sequence, 15fps, 544x960
12
+
13
+ STABLE DIFFUSION
14
+
15
+ 2. Use an anime consistent model and LorA
16
+
17
+ 3. t2i : Generate the reference picture with one of the first depth frame
18
+
19
+ 4. i2i : Using Multi-Control Net a. Batch depth with no pre-processor b. Reference with the reference pic generated in 2. c. TemporalKit starting with the reference pic generated in 2.
20
+
21
+ POST PROCESS
22
+
23
+ 5. FILM interpolation (x2 frames)
24
+
25
+ 6. Optionnal : Upscale x2 (Anime6B)
26
+
27
+ 7. FFMPEG to build the video (30fps)
28
+
29
+ 8. Optionnal : Deflicker with Adobe
30
+
31
+ NB :
32
+
33
+ Well known animes are usually rendered at low fps, so I wouldn't overkill it at 60fps to keep the same anime feeling (+ it would take ages to process each step, and also randomly supported by socials apps like TikTok)
34
+
35
+ Short hair + tight clothes are our friends
36
+
37
+ Good consistency even without Deflicker
38
+
39
+ Depth is better than Openpose to keep hair/clothes physics
40
+
41
+ TO IMPROVE :
42
+
43
+ - Hands gestures are still awful even with the TI negatives
44
+
45
+ - Background consistency by processing the character separately and efficiently
46
+
47
+ Hope you enjoy it, all this gives me new ideas...